Page 53 - Hydrocarbon Exploration and Production Second Edition

40 Exploration Methods and Techniques

typically with a spacing between 12.5 and 50 m. The processed result is a 3D
‘volume’ or cube of data (Figure 3.22) that can be viewed along all three axes (line,
trace, time/depth). These days the volumes can also be sliced along an ‘arbitrary
line’ such as along the axis of a meandering channel. A 3D seismic volume typically
contains thousands of traces. It is clear to see that in the course of the processing
phase, such large volumes require huge amounts of disk space.

3.2.3. Seismic interpretation
After processing has been completed, the data are loaded onto a workstation for
interpretation by geologists and geophysicists. The workstations are powerful
computers, often Linux-based with dual screen capacity to allow the interpreter to
look at the data in vertical section on one screen and in map view on the other. The
first step in the interpretation cycle is to ‘tie’ the seismic data to existing well data in
order to identify what the important reflector events correspond to, for example top
of the reservoir or top of the main seal. In a mature field there are typically dozens
of wells to calibrate to, but in exploration areas there may only be a couple,
sometimes located several kilometres away.
The main reflectors or horizons are digitised from the screen (picked) and stored
in a database; the same is done for the faults (Figure 3.23). In this way the structure
of the field is mapped out (Figure 3.24) and potential structural or stratigraphic traps
are delineated. More detailed analysis can lead to identification of the internal
architecture of the reservoir interval, such as separate sand bodies within a complex
channel system.
Nowadays geoscientists and engineers prefer to view seismic data not in terms of
reflection data with the characteristic wavelet signature, but in terms of acoustic
impedance. This is achieved by seismic inversion, a process which removes the
influence of the wavelet and represents the data in a geologically meaningful way,
namely as a function of rock properties. Inversion requires careful calibration to well
data and knowledge of the broad geological model of the subsurface.
Once the interpretation has been completed in the time domain, the interpreted
surfaces need to be converted to depth for use in the geological and engineering
model. Depth conversion again requires knowledge of the seismic velocity and any
significant variations, both lateral and vertical, that may be present. There are several
methods of depth conversion. A simple method is to derive seismic interval
velocities for a number of key intervals and then to calculate the thickness for each
interval before summing them. This method is called ‘isochoring’ and gives a
reasonable result in areas not affected by velocity variations. Another method is to
build a velocity model based on stacking velocities. In areas of complex geology,
more intricate methods are required and even then there can be large discrepancies
between true depth and calculated depths.

3.2.4. Seismic attributes

The development of post-stack processing algorithms has allowed 3D seismic data to
be interrogated in increasingly sophisticated ways. Structural attributes of the data